minimum gpu memory requird and runtime error while running YOLOs, YOLOm, YOLOl

[SumayyaInayat]

Hi,
Hope you are doing well.
Can you please tell me how much gpu ram is required to reproduce your results with YOLOx?
And upon running YOLOs or YOLOm it gives dimension miss match error, Can you please help me with this?

RuntimeError: Given groups=1, weight of size [1280,1280,3,3], expected input[1,512,4,4] to have 1280 channels, but got 512 channels instead

Thanks!!!

[Yipeng Gao]

Hello, I ran all the experiments on V100 GPU, which has 32G memory :(

You can run results on YOLOs or YOLOm for comparisons. The key lies in the "Ranker" class here, which maps the feature map to embeddings for alignment. I specify the input dim as 1280, producing bugs when you use YOLOs or YOLOm. You can modify this hyperparameter or make it as an input argument.

Feel free to ask me if any other questions exist :)))

[SumayyaInayat]

Hi,

Oh ok so I was trying it on max 12 GB gpu with batch 2 but it gave Cuda out of memory error.
I will make the required changes in Ranker class and then ask for any help if I want not able to solve it myself.

Thanks alott for helping!!